Data sets generally include a plurality of items which can be configured as data records. In order to determine a particular item from the plurality of items, search strategies can be used which orient themselves by a classification of the data set.
However, in this case it is required that the user who wants to perform the search is aware of the type of classification. Usually, the following classifications can be present:                A flat classification characterized by a simple list of features.        A hierarchical classification which is based on a tree structure. A tree provided for this purpose is branched at nodes with a relation existing between the nodes and branches.        A network classification in which the individual items are interconnected by cross-references, wherein each item can theoretically be linked with each other item. The handling of such a network classification is generally problematic.        A faceted classification has a star-shaped topology, wherein a center is linked with a plurality of peripheral points.        
Depending upon a classification implemented in the respective data set, the user can filter out items of the set in a search using different strategies. However, this procedure is difficult if the user does not know a respective classification mentioned above.
According to the prior art, it is also possible to sense a search term within a data set in the framework of a full-text search. However, if there is no taxonomy for the data set, such a full-text search is performed in an unspecific way.
A database is known which includes a list of all female and male Nobel Prize laureates. Here, to each female or male Nobel Prize laureate, are assigned among others the nationality, the affiliation to an organization, the category in which the Nobel Prize was won, and the year in which the Nobel Prize was awarded. For these mentioned categories individual properties can be retrieved, so that a female or male Nobel Prize laureate can be searched for under consideration of the nationality, for example. However, the categories of this database are rigidly predetermined and can not be expanded.
On the basis of these facts, the above mentioned methods, the device and the computer program product having the features of the respective independent and accessory claims are proposed.